Visokorazpoložljvi sistem za nadzor in opazovanje

ZUPAN, NEJC (author), Brodnik, Andrej (mentor)

S porastom uporabe spletnih tehnologij in storitev je nadzor omrežij postal nuja. Vse več storitev je danes na voljo preko spleta, s čimer se je povečalo tudi število uporabnikov. Ker želimo storitvam zagotoviti čim boljšo stabilnost, se poslužujemo visokorazpoložljivih in nadzornih sistemov. V diplomskem delu smo obravnavali nekaj nadzornih protokolov in sistemov za nadzor omrežij, na podlagi katerih smo izbrali najprimernejše in jih podrobneje opisali. Temu opisu sledi eksperimentalni del, v katerem smo postavili prototip visokorazpoložljivega nadzornega sistema, ki tudi v primeru preklopa uporabniku nudi nemoteno delovanje. Naša rešitev vsebuje nadzorni sistem Zabbix, ki z agenti komunicira preko izbranega SNMP protokola. Za nastavitev preklopa smo uporabili odprtokodne programe Pacemaker, Corosync in Heartbeat, za serviranje spletnega vmesnika pa Nginx. Podatki se hranijo v podatkovni bazi PostgreSQL, ki deluje v aktivno-pasivnem načinu in prav tako podpira preklop med vozlišči.

• Language: Slovenian
• Keywords: visokorazpoložljivi sistemi, Zabbix, preklop, SNMP, nadzorni sistem
• Work type: Bachelor thesis/paper
• Organization: FRI - Faculty of Computer and Information Science
• Year: 2014
• PID: 20.500.12556/RUL-29524
• Publication date in RUL: 19.09.2014

Secondary language

• Language: English
• Title: High availability system for control and monitoring

Abstract

As we see an increase in the use of web services, having a control over networks became a priority. More and more services are available online and consequently, the number of users has increased. Since our aim is to get more stability, we are using high available monitoring system. In this thesis we addressed few protocols and systems for monitoring networks, based on which we chose the most appropriate ones and described them in a greater detail. Following that description is experimental part in which we set up a high available monitoring system that ensures smooth user interaction even during failover to the other node. The solution includes monitoring system Zabbix which communicates with the agents via the selected SNMP protocol. For failover we used open-source programs like Pacemaker, Corosync and Heartbeat, while we used Nginx for monitoring web page. Data are stored in the PostgreSQL database that operates in active-passive mode and also supports failover between nodes.

• Keywords: high availability, Zabbix, failover, SNMP, monitoring system